Chemoresistance to nucleoside analog drugs (e.g., gemcitabine) is one of the main underlying reasons for the extremely poor prognostic state of pancreatic cancer. While the pivotal roles of cellular nucleoside trans- porters (NTs) and intercellular gap junctional connexins (Cxs) in determining tumor exposure of nucleoside drugs have begun to emerge, their functional interplay in determining nucleoside analog chemosensitivity and their potential in predicting response to chemotherapy remain unknown. Our long-term goal is to improve the chemotherapeutic management of pancreatic cancer. The overall objective of this R15 application is to investi- gate NT-Cx interplay in pancreatic cancer chemotherapy. The central hypothesis is that the type of NT-Cx combination expressed will dictate chemosensitivity and that the specific combination(s) can then be targeted to increase anti-tumor efficacy. This hypothesis has been formulated on the basis of the continuum of studies on NTs and Cxs and the recent discovery of non-genetic functional alterations in NT and Cx subtypes in pan- creatic tumors. The rationale underlying the proposed research is that understanding the precise determinants of nucleoside analog sensitivity will help in the preselection of patients suitable for this type of therapy and the identification of clinical strategies to increase efficacy in poor responders. Specific aim 1 will define the role of NT-Cx combinations in nucleoside analog chemosensitivity. The working hypothesis is that gemcitabine cyto- toxicity will be superior in cells expressing the concentrative NT 1 (hCNT1)-Cx32 combination, and that chemo- resistance due to the loss of one component can be effectively compensated by the other. This hypothesis is based on the expression and permeation characteristics of NTs and Cxs in pancreatic tumors. Specific aim 2 will determine the regulators governing NT-Cx interplay to improve chemosensitivity. The working hypothesis is that the favorable manipulation of NT-Cx combinations to increase drug exposure can be achieved by manipu- lating cadherins or microRNAs. This hypothesis is based on the applicant's studies showing cadherin control of Cx assembly and microRNAs as putative regulators of hCNT1. Specific aim 3 will investigate the in vivo pre- dictability of nucleoside analog response using NT-Cx combinations. The working hypothesis is that a compo- site index comprising of specific NTs and Cxs will better predict in vivo gemcitabine response than with the ex- isting single index (i.e., equilibrative NT 1 (hENT1) alone). This hypothesis is based on the extrapolation of pre- liminary in vitro results obtained in cultured pancreatic cancer cells. The anticipated outcomes of this work are the delineation of NT-Cx interplay in nucleoside analog chemosensitivity, strategies to improve NT-Cx- mediated drug targeting, and evaluation of clinical measures for predicting chemotherapeutic responses in pancreatic cancer subtypes. This approach is innovative because it focuses on a comprehensive index for judging chemotherapeutic response. This contribution is significant because it will enable subsequent transla- tional studies that are expected to improve treatment and survival outcomes in pancreatic cancer patients.